coffee/libyuv
1
0
Fork 0
mirror of https://chromium.googlesource.com/libyuv/libyuv synced 2025-12-07 01:06:46 +08:00
Code Issues Packages Projects Releases Wiki Activity

Align branch targets for NaCL validator.

Browse Source 
BUG=253
TEST=validator
R=johannkoenig@google.com, ryanpetrie@google.com

Review URL: https://webrtc-codereview.appspot.com/2033004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@758 16f28f9a-4ce2-e073-06de-1de4eb20be90
This commit is contained in:
fbarchard@google.com 2013-08-14 17:58:33 +00:00
parent c4a70492c0
commit d459f382fd
4 changed files with 23 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
README.chromium
View File

@ -1,6 +1,6 @@
Name: libyuv Name: libyuv
URL: http://code.google.com/p/libyuv/ URL: http://code.google.com/p/libyuv/
Version: 757 Version: 758
License: BSD License: BSD
License File: LICENSE License File: LICENSE

2
include/libyuv/version.h
View File

@ -11,6 +11,6 @@
#ifndef INCLUDE_LIBYUV_VERSION_H_ // NOLINT #ifndef INCLUDE_LIBYUV_VERSION_H_ // NOLINT
#define INCLUDE_LIBYUV_VERSION_H_ #define INCLUDE_LIBYUV_VERSION_H_
#define LIBYUV_VERSION 757 #define LIBYUV_VERSION 758
#endif // INCLUDE_LIBYUV_VERSION_H_ NOLINT #endif // INCLUDE_LIBYUV_VERSION_H_ NOLINT

4
source/scale.cc
View File

@ -781,6 +781,8 @@ static void ScaleAddRows_SSE2(const uint8* src_ptr, ptrdiff_t src_stride,
paddusw xmm1, xmm3 paddusw xmm1, xmm3
sub ebp, 1 sub ebp, 1
jg yloop jg yloop
align 16
ydone: ydone:
movdqa [edi], xmm0 movdqa [edi], xmm0
movdqa [edi + 16], xmm1 movdqa [edi + 16], xmm1
@ -1358,6 +1360,7 @@ static void ScaleAddRows_SSE2(const uint8* src_ptr, ptrdiff_t src_stride,
"mov %5,%2 \n" "mov %5,%2 \n"
"test %2,%2 \n" "test %2,%2 \n"
"je 3f \n" "je 3f \n"
".p2align 4 \n"
"2: \n" "2: \n"
"movdqa (%0),%%xmm2 \n" "movdqa (%0),%%xmm2 \n"
"add %6,%0 \n" "add %6,%0 \n"
@ -1368,6 +1371,7 @@ static void ScaleAddRows_SSE2(const uint8* src_ptr, ptrdiff_t src_stride,
"paddusw %%xmm3,%%xmm1 \n" "paddusw %%xmm3,%%xmm1 \n"
"sub $0x1,%2 \n" "sub $0x1,%2 \n"
"jg 2b \n" "jg 2b \n"
".p2align 4 \n"
"3: \n" "3: \n"
"movdqa %%xmm0,(%1) \n" "movdqa %%xmm0,(%1) \n"
"movdqa %%xmm1,0x10(%1) \n" "movdqa %%xmm1,0x10(%1) \n"

18
source/scale_argb.cc
View File

@ -236,6 +236,8 @@ static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb,
lea edi, [edi + 8] lea edi, [edi + 8]
sub ecx, 2 // 2 pixels sub ecx, 2 // 2 pixels
jge xloop2 jge xloop2
align 16
xloop29: xloop29:
add ecx, 2 - 1 add ecx, 2 - 1
@ -244,6 +246,8 @@ static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb,
// 1 pixel remainder // 1 pixel remainder
movd xmm0, qword ptr [esi + eax * 4] // 1 source x0 pixels movd xmm0, qword ptr [esi + eax * 4] // 1 source x0 pixels
movd [edi], xmm0 movd [edi], xmm0
align 16
xloop99: xloop99:
pop edi pop edi
@ -313,6 +317,8 @@ static void ScaleARGBFilterCols_SSSE3(uint8* dst_argb, const uint8* src_argb,
lea edi, [edi + 8] lea edi, [edi + 8]
sub ecx, 2 // 2 pixels sub ecx, 2 // 2 pixels
jge xloop2 jge xloop2
align 16
xloop29: xloop29:
add ecx, 2 - 1 add ecx, 2 - 1
@ -328,6 +334,8 @@ static void ScaleARGBFilterCols_SSSE3(uint8* dst_argb, const uint8* src_argb,
psrlw xmm0, 7 psrlw xmm0, 7
packuswb xmm0, xmm0 // argb 8 bits, 1 pixel. packuswb xmm0, xmm0 // argb 8 bits, 1 pixel.
movd [edi], xmm0 movd [edi], xmm0
align 16
xloop99: xloop99:
pop edi pop edi
@ -485,6 +493,8 @@ static void ScaleARGBRowDownEvenBox_SSE2(const uint8* src_argb,
} }
#define HAS_SCALEARGBCOLS_SSE2 #define HAS_SCALEARGBCOLS_SSE2
// TODO(fbarchard): p2align 5 is for nacl branch targets. Reduce using
// pseudoop, bundle or macro.
static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb, static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb,
int dst_width, int x, int dx) { int dst_width, int x, int dx) {
intptr_t x0 = 0, x1 = 0; intptr_t x0 = 0, x1 = 0;
@ -501,7 +511,7 @@ static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb,
"paddd %%xmm3,%%xmm3 \n" "paddd %%xmm3,%%xmm3 \n"
"pextrw $0x3,%%xmm2,%k4 \n" "pextrw $0x3,%%xmm2,%k4 \n"
".p2align 4 \n" ".p2align 5 \n"
"2: \n" "2: \n"
"paddd %%xmm3,%%xmm2 \n" "paddd %%xmm3,%%xmm2 \n"
"movd (%1,%3,4),%%xmm0 \n" "movd (%1,%3,4),%%xmm0 \n"
@ -514,11 +524,14 @@ static void ScaleARGBCols_SSE2(uint8* dst_argb, const uint8* src_argb,
"sub $0x2,%2 \n" "sub $0x2,%2 \n"
"jge 2b \n" "jge 2b \n"
".p2align 5 \n"
"29: \n" "29: \n"
"add $0x1,%2 \n" "add $0x1,%2 \n"
"jl 99f \n" "jl 99f \n"
"movd (%1,%3,4),%%xmm0 \n" "movd (%1,%3,4),%%xmm0 \n"
"movd %%xmm0,(%0) \n" "movd %%xmm0,(%0) \n"
".p2align 5 \n"
"99: \n" "99: \n"
: "+r"(dst_argb), // %0 : "+r"(dst_argb), // %0
"+r"(src_argb), // %1 "+r"(src_argb), // %1
@ -593,6 +606,7 @@ static void ScaleARGBFilterCols_SSSE3(uint8* dst_argb, const uint8* src_argb,
"sub $0x2,%2 \n" "sub $0x2,%2 \n"
"jge 2b \n" "jge 2b \n"
".p2align 4 \n"
"29: \n" "29: \n"
"add $0x1,%2 \n" "add $0x1,%2 \n"
"jl 99f \n" "jl 99f \n"
@ -605,6 +619,8 @@ static void ScaleARGBFilterCols_SSSE3(uint8* dst_argb, const uint8* src_argb,
"psrlw $0x7,%%xmm0 \n" "psrlw $0x7,%%xmm0 \n"
"packuswb %%xmm0,%%xmm0 \n" "packuswb %%xmm0,%%xmm0 \n"
"movd %%xmm0,(%0) \n" "movd %%xmm0,(%0) \n"
".p2align 4 \n"
"99: \n" "99: \n"
: "+r"(dst_argb), // %0 : "+r"(dst_argb), // %0
"+r"(src_argb), // %1 "+r"(src_argb), // %1